The U.S. stock market's tech sector has made a strong comeback! The earnings season has reignited confidence in AI, with the market value of the NASDAQ 100 surging by $3.5 trillion in just four days.

After a strong earnings report boosted investor confidence in the prospects of artificial intelligence (AI), U.S. tech stocks saw a dramatic reversal, driving the market capitalization of the Nasdaq 100 index up by $3.5 trillion in just four trading days. Data shows that over the past four trading days, this benchmark tech index has risen by 9.3%, marking its largest increase since April 2025when the market was recovering from the turbulence caused by what former U.S. President Trump referred to as "Liberation Day" tariffs. Nasdaq 100 Index records one of its largest four-day increases since Liberation Day in 2025. The much better-than-expected performance during the second quarter earnings season for U.S. stocks has rekindled investor belief that massive AI investments will not only persist but are already starting to yield returns for some industry giants. Data indicate that the cloud backlogs for hyperscale cloud service providers surged over 150% year-on-year, totaling about $1.7 trillion, with growth rates far surpassing the approximately 80% increase in capital expenditures during the same period. JPMorgan pointed out that this significant disparity suggests that the potential revenue returns from AI infrastructure investments are exceeding market expectations, indicating that the valuation digestion pressures on tech giants may be nearing resolution. Furthermore, during this earnings season, tech stocks have a 90% probability of exceeding profit expectations even under high expectations, while analysts continue to raise their earnings forecasts. However, with the upward revision in earnings expectations, tech stock valuations have experienced a noticeable compressionfollowing the market adjustment in July, the forward price-to-earnings ratio for the S&P 500 information technology sector has dropped to about 20 times, approaching the lowest level in the past year and sitting at the first percentile of historical valuation ranges, below the average level of about 23 times over the past decade. In this regard, JPMorgan noted that the forward price-to-earnings ratio for large-cap tech stocks (excluding semiconductors) is currently more than two standard deviations below the historical average since 2018. If valuations were to recover to a level one standard deviation below the historical average, there could be approximately 30% upside potential; if restored to the vicinity of long-term averages, the potential upside could reach about 56%. Over the past year, the speed of profit expectation increases has been double that of stock price increases. This rebound also marks a quick counterattack by the Nasdaq 100 index. More importantly, the scope of the tech sector's rebound is broad, with semiconductor firms, software companies, and hyperscale cloud computing companies all seeing gains. Among the standout companies, over the past four trading days, SanDisk (SNDK.US) has risen 41%, Palantir (PLTR.US) has increased by 32%, Microsoft Corporation (MSFT.US) has gained 26%, and both Alphabet Inc. Class C (GOOGL.US) and NVIDIA Corporation (NVDA.US) have risen by 11%. David Reinvil, the head of the Sycomore Sustainable Tech fund, stated, The trading approach that involved shorting capital expenditure firmsnamely, hyperscale cloud computing companieswhile going long on beneficiaries of capital expenditure has ended, and I view this as a good thing. The market is no longer a simple binary trade. Meanwhile, during last month's correction in tech stocks, a large-scale deleveraging occurred that led hedge funds and other fast trading funds to close out a significant number of short positions. Currently, these funds have started to flow back in and buy into the tech sector. According to data from the Prime Brokerage unit of Goldman Sachs Group, Inc., last week saw hedge funds increasing their holdings in the information technology sector at the fastest rate since December 2022. Goldman Sachs Group, Inc. teams indicated that the "Magnificent Seven" overall received net purchases, but the overall holding levels remain relatively low, suggesting there is still room for further position increases going forward. However, the latest earnings season has also highlighted a clear divergence among tech giants. Investors are rewarding companies that can demonstrate real AI revenue growth, while penalizing those with more uncertain returns on AI investments. Roland Caloyan, a strategist at France's Industrial Bank, remarked, The new changes in AI trading show significant differentiation within semiconductor companies and hyperscale cloud computing firms. This means that portfolio managers who cannot invest in indexes and must actively select individual stocks are now facing greater challenges in stock picking. The disparity in performance is quite marked. Between the close on July 29 and the close on August 4, Nebius (NBIS.US) saw its stock price increase by 52%, while Apple Inc. (AAPL.US) saw a decline of 8.5%. Within the broadly AI-related tech stock sector, there was a remarkable 61-point gap between the best and worst performers. For strategists like Parag Tatti at Deutsche Bank Aktiengesellschaft, the trend of capital flowing out of large-cap tech stocks reached a bottom last week. Following this, the market positions in these stocks began to rise moderately, indicating that there remains further upside potential. Deutsche Bank strategists believe that within the tech sector, the risk-reward ratio for hyperscale cloud computing firms is the best. Currently, these companies performance relative to the S&P 500 index has just recovered from the weakest level in the last three years. The strategists stated, We believe that the trend of capital flowing back into the tech sector still has room for further development, and according to historical patterns, tech stocks typically achieve about a 20% outperformance. They pointed out, This will mark the fifth rotation in the tech sector over the past three years, with market focus repeatedly switching between high growth opportunities and bubble risks, and this switching pace is accelerating. JPMorgan believes that if the market narrative around AI shifts from Is capital expenditure excessive? to Investment returns are being realized, the next phase of tech stock rises may come more from internal sector rotation rather than solely relying on semiconductor stocks continuing to rise. From a technical perspective, the U.S. technology seven giants index (MAGS) has rebounded nearly 10% from recent lows, regaining its position above the 200-day moving average and nearing the long-term upward trend line since last April. JPMorgan believes that the current 200-day moving average has leveled off, indicating the market is experiencing a prolonged consolidation phase. Historical experience shows that the longer the sideways time, the stronger the breakout momentum after the direction is chosen.